Erosive wear by solid particle impingement was tested for 13 kinds of commercial Al₂O₃ ceramics. The erosion rate of Al₂O₃ ceramics is inversely proportional to the specific gravity, Al₂O₃ content and Vickers hardness of the target ceramics. The erosion rates of high density Al₂O₃ ceramics composed of more than 99wt% of Al₂O₃ are 3.9-4.5×10^-3cm³/cm³, the same order as those of gas pressure sintered and hot pressed Si₃N₄. The erosion rate (V) of Al₂O₃ ceramics is related to its hardness (H) and fracture toughness (K_IC) by the equation V=0.56H^-1.2K_IC^-0.94, derived from the multiple regression analysis. Indices for H and K_IC in the equation are different from those for hot pressed Si₃N₄, -0.27 for H and=7.1 for K_IC. These differences are caused by the differences in erosion mechanism.

View full abstract

We proposed an atomic configurational model on an asymmetric boundary between 3C (111) and 6H (1012) faces of SiC single crystals. In this model, the boundary energy was estimated with the numbers of dangling bonds at the boundary with different energies on carbon or silicon atom. The results were extended to the boundary between SiC 6H (0001) and 6H (1012) faces for comparing with the experimental results of high-resolution electron microscopy. It is concluded that the geometrical model is useful for studying the boundary between c-face and others of SiC crystals.

View full abstract

The non-isothermal crystallization kinetics in ZrF₄-BaF₂-AlF₃ and ZrF₄-BaF₂-LaF₃ glasses were analyzed. It was found that the crystallization temperatures increase as the contents of AlF₃ and LaF₃ increase. Al and La atoms retarded the crystal growth in these glasses. The crystal phases precipitated in ZrF₄-BaF₂-AlF₃ glasses were β-BaZrF₆ and β-BaZr₂F₁₀ crystals, and the volume fraction of β-BaZr₂F₁₀ crystal increased as the content of AlF₃ increased. The crystal phases in ZrF₄-BaF₂-LaF₃ glass were β-BaZrF₆ and an unknown phase. The activation energies for the crystal growth in these glasses were lower than those in ZrF₄-BaF₂ glasses. The mechanism of crystal growth was different from that in ZrF₄-BaF₂ glasses.

View full abstract

Fick's 2nd law of diffusion was solved for small values of time (the solution of error function) for the problem of diffusion in a sphere which contains a surface layer where the diffusion coefficient is higher than that in the interior of the sphere. A solution to the same problem in the conduction of heat was already obtained by Carslaw by separation of variables. The solution for the heat conduction was rewritten as the solution for the diffusion problem, and then these two types of solutions were compared with each other. There was very good agreement between them. The range of variables where the solution of error function is applicable was discussed.

View full abstract

Effects of various additives on densification, thermal conductivity and electrical resistivity of SiC ceramics were studied. Addition of elemental Be, B and Al or their compounds gave higher-than-98% theoretical densities by hot-pressing at 2040°C for 1h at a load of 20MPa. Thermal conductivity and electrical resistivity of dense SiC ceramics were measured at 25°C. They depend largely on the additives. SiC ceramics containing Be or its compounds had high thermal conductivity of above 260W/m·K and electrical resistivity of above 1×10¹³Ω·cm. On the other hand, SiC ceramics containing B or its compounds had thermal conductivity of 110-170W/m·K and electrical resistivity of 10⁴-10¹¹Ω·cm. Al or its compounds gave the lowest thermal conductivity and electirical resistivity of 60-80W/m·K and 0.6-27Ω·cm, respectively. Thermal conductivity and electrical resistivity of SiC ceramics depend on the amount of impurities and free carrier concentration in the SiC grains. The effects of additives on thermal conductivity and electrical resistivity of SiC ceramics examined were attributed to the different solubilities of the additives into SiC grains. High thermal conductivity and high electrical resistivity of SiC ceramics containing Be or its compounds were shown to be caused by low solubility of Be into SiC grains.

View full abstract

OMCVD of TiO₂-SnO₂ mixed films using titanium isopropoxide and tetrabutyltin as starting materials was investigated. In the temperature region between 500° and 550°C crystalline films with desired composition were obtained. Pure titania films had the anatase crystal form, but the doped ones with more than 25% SnO₂ exhibited the rutile form. The size of crystallites in as-deposited films depended on the film composition and was found to be smallest near Ti/Sn ratio of 1:1. When the films were annealed at 1000°C over 50h, the crystalline size increased, however, its composition dependence was the same as that in as-deposited films. From this tendency and the composition dependence of lattice constants of the annealed films, it is concluded that the phase separation takes place through the spinodal decomposition during heat treatment. The as-deposited mixed films had no photo-conductivities but had moderate dark conductivities which strongly depended on the ambient gas components. Thus they are applicable to a basic material for gas sensors.

View full abstract

Fracture behavior of 15vol% SiC fiber (SCS-6 Φ 140μm; AVCO Specialty Materials Div., U.S.A) reinforced Si₃N₄ composite materials was directly observed with an in-situ observation apparatus composed of bending machine and scanning electron microscope. When load is applied to a notched specimen, a crack initiates from the bottom of the notch and propagates in the specimen non-catastrophically, because the crack propagation is stopped successively at the fibers. With further increase of the load, the crack elongates and halts over again and again, and finally propagates along the fibers. Fatigue crack growth was not detected after 100 times of cyclic load application with a constant strain rate and 300 times application of impact load, when the maximum applied stress of the cyclic loading is less than 75% of the maximum bending strength of notched specimens. In the cyclic loading process, moreover, it was observed that the crack width becomes wide and narrow on loading and unloading, respectively.

View full abstract

A refinement technique of X-ray diffraction profiles was applied to the quantitative analysis of a preferentially oriented particle with a layer structure in the plane of a sample-holder. This approach seems to be superior to the previous methods where special techniques of sample preparation are required to reduce the effects of preferred orientation. The present approach suggested that the effects of preferred orientation could be cancelled out by correcting the observed intensity for the various reflections using a predetermined orientation distribution function of plate-like samples. To study the influence of stacking disorder in minerals on the quantitative analysis, a fundamental equation in X-ray powder diffraction method was also developed on the basis of the one dimensional stacking disorder theory. To make the quantitative check of the present method, two samples were prepared; muscovite and quartz. They were crushed and sieved into different particle size ranges, such as -2, 3-5, 5-10, 2-10 and 20-40μm. In the two component systems with various mixing ratios, the validity of the present method was confirmed with refinement of the entire X-ray intensity profiles for every admixture in which each mineral was weighed earlier. The discrepancy between the calculated and prepared mineral compositions was within 6 weight percent for all mixtures.

View full abstract

Clay minerals in the Hadong Kaolin, Korea were characterized by X-ray diffraction, transmission electron microscopy and chemical analysis. Two kinds of samples, pink DK-7 purified with a high gradient magnetic separator and white WA were separated to <5, <2 and <0.3μm fractions. Clay minerals contained in DK-7 and WA samples were halloysite (10Å), kaolinite and a small amount of sericite. The amount of kaolinite in DK-7 was higher than that in WA. Kaolinite and sericite were not found in the <0.3μm fraction samples. Particles of halloysite (10Å) were a short tube and those of kaolinite and sericite were irregular plates. Chemical composition of halloysite (10Å) in the <5μm fraction sample showed fairly Al₂O₃-rich composition than ideal one due to the co-existence of low crystalline aluminum hydroxide. The amount of iron in halloysite was around 1wt% and 0.5wt% in DK-7 and WA, respectively.

View full abstract

A whiteware body composition containing 30% alumina was investigated to study the effects of milling and sieving raw materials on bending strength and Weibull distribution of fired bodies. Test bars were formed by slip casting into a rectangular shape (10b×5b×70mm) and fired at 1350°C for 1h in a reducing atmosphere. Mean bending strength of the fired specimens increased with milling time, almost proportionally to time. Weibull modulus, however, decreased with milling time, because of a small amount of coarse grains which remained in the slip even after long milling. Removing the coarse grains in the slip through a finer sieve resulted in increased strength and higher Weibull modulus. It was found that quartz grains included in the raw feldspar were more effective as fracture sources than feldspar itself. A mean bending strength of 240MPa and Weibull modulus of 21 were obtained from the slip milled for 24h and sieved through 350mesh.

View full abstract

The thermal conductivity of alumino-silicate fibrous insulators with various bulk densities was measured by the transient hot wire method under various reduced pressures in order to propose a simple equation for predicting their thermal conductivities. The results obtained were summarized as follows:
(1) The thermal conductivity of ceramic fibrous insulators with various bulk densities decreased gradually with decreasing pressure ranging from 760 to 10mmHg and converged to an extremely low constant value under 10^-2mmHg after a sharp decrease between 10 and 10^-1mmHg, due to the pressure dependence of thermal conductivity of air. Their thermal conductivity increased with increasing bulk density all over the pressure range covered in the thermal conductivity measurement.
(2) The thermal conductivity of ceramic fibrous insulators in helium, oxygen and freon-12 gas atmospheres also increased with an increase in bulk density, and their thermal conductivity in a gas atmosphere having high thermal conductivity such as helium was higher than those in gases with lower thermal conductivity such as freon-12.
(3) In alumino-silicate fibrous insulators, the following experimental relation among λ (thermal conductivity of insulator), λ_s (that of fiber: 1.07W/m·K), λ_f (that of gas being filled between fibers) and φ (volume fraction of fiber) was obtained.
λ=aλ_s+bλ_f+(1-a-b)λ_sλ_f/(1-y)λ_s+yλ_f
a=0.169φ^1.57
b=(1-φ)^1.16
y=0.81-0.043lnφ
This equation enabled us to calculate the thermal conductivity of the fibrous insulators in various gas atmospheres (-0.2W/m·K) at room temperature within the error of about ±10%.

View full abstract

The purpose of this paper is to investigate the applicability of the ball-on-disk method to ceramics, which is considered to be suitable for measurement of a very small amount of wear. The bearing steel ball (SUJ2) was slid on ceramic disks of SiC, Si₃N₄, Al₂O₃ and PSZ, and the friction and wear properties were measured. The wear volume was calculated from the dimensions of the wear track on the ball. The friction force and wear volume remained almost constant for the sliding velocity region of 10-20cm/s, at 1kg load and after 8000total rotation. Then, by changing the total rotation number at 17.5cm/s sliding velocity, the wear volume was found to be proportional to the weight loss of the ball, and also to the sliding distance. Those results showed that the coefficient of friction and the specific wear rate calculated from these data are reliable. The specific wear rate of SUJ2 ball was lowest on the PSZ disk, followed by SiC and then Si₃N₄. The wear on Al₂O₃ disk was very large. The wear of the ceramic disk itself was estimated from the cross sectional profile of the wear circle. The Si₃N₄ disk showed highest wear, followed by SiC and then PSZ. The wear of Al₂O₃ disk was hardly detected.

View full abstract

Titanium nitride and tantalum nitride powders were synthesized by nitriding metal powders under high nitrogen pressures of 1 to 10MPa. Starting metal powders were diluted with corresponding nitride powders to control the reaction temperature and prevent agglomeration of molten particles. The characteristics of combustion reaction in this process were investigated by measuring the propagating velocity of the combustion front for packed reactants with various dilution levels. The nitriding process in combustion reaction was explained as follows; in the combustion synthesis of TiN, nitride layer was formed on molten particles in the early stage of combustion reaction, which prevented coalescing of molten particles. On the other hand, in TaN the combustion proceeded by solid state reaction between metal particles and nitrogen gas. In both cases diffusion of nitrogen atom through the nitride was considered to be a rate-determining step in the combustion reaction.

View full abstract

A BaTiO₃ precursor sol was prepared by controlled hydrolysis of Ba-Ti mixed alkoxide solution in methoxyethanol. SrTiO₃, BaZrO₃ and SrZrO₃ precursor sols were also synthesized by the similar method. The hydrolysis condition of Ba-Ti system was as follows; a water-methoxyethanol mixed solution with less than 50vol% of water was added to 3.6×10^-4mol/ml of alkoxide solution at room temperature. On adding over 50vol% of water, partial gelation occurred, and heterogeneous sol or gel was formed. For Sr-Ti system, the hydrolysis condition was the same as that for Ba-Ti system. For Ba-Zr and Sr-Zr system, the concentration of water which produces partial gelation was over 66vol%. The formed precursor sols transformed to one phase BaTiO₃, SrTiO₃, BaZrO₃ or SrZrO₃ by calcination. Furthermore, the result of EDX analysis for SrTiO₃ showed that the obtained powder had homogeneous composition and its homogeneity was kept up to 1000°C. Similar results may be expected for other systems.

View full abstract

The gas pressure sintering of silicon nitride with SiC containing Al₂O₃ and CeO₂ as densification aid was investigated from 1800°C and 2MPa to 2050°C and 6MPa of nitrogen. The Si₃N₄-20wt% SiC-5wt% Al₂O₃-5wt% CeO₂ composition sintered at 2050°C by two-step method had a density of 3.24g/cm³, flexural strength of about 780MPa, fracture toughness of about 4.9MN·m^-3/2 and Vickers hardness of about 1700kg/mm². It was found that addition of SiC to Si₃N₄ suppressed the growth of Si₃N₄ grains effectively.

View full abstract

In order to develop a joining technique for reaction bonded Si₃N₄ which ensures good high-temperature strength of joined body, Al foil was used as an insert material. The joining was conducted in 1atm N₂ atmosphere at temperatures from 1470 to 1970K, under pressures up to 20MPa for 1800s. Direct joining without any insert material was also carried out for comparison. The microstructure of joined regions was observed with a high resolution electron microscope, and the strength of the interface was measured by 3-point bending test as a function of joining temperature and joining pressure. The results obtained are summarized as follows.
(1) Direct joining is almost impossible under the above conditions.
(2) When joined by using Al foil at 1970K under 20MPa, the bending strength at 1670K attains to 154MPa.
(3) The interface layer is composed of unreacted Al and AlN-polytype Sialon, and Al joins to Si₃N₄ matrix with the Sialon layer in-between. As the joining temperature rises, the amount of Sialon increases at the expense of Al and Al is completely converted to Sialon at 1970K.
(4) Along grain boundaries of Sialon, there exists a non-crystalline phase of 1-3nm in thickness. The non-crystalline phase can be an extended grain boundary.
(5) When joined at 20MPa, the bending strength is insensitive to the joining temperature at room temperature but very sensitive at 1670K. This is explained by, (a) the amount of unreacted Al and its degree of adhesion at the interface, and (b) the residual stress supposedly induced by the difference in thermal expansion coefficient between the Sialon and the Si₃N₄ matrix.
(6) The strength increases with increasing joining pressure, but saturates above about 10MPa. This is explained by the joining-pressure dependence of the amount of the Sialon formed.

View full abstract

BaTiO₃ ceramics including 0.2mol% Ta₂O₅, increases its compactness due to the agglomeration during the first sintering period, but the compactness decreases with decreasing the particle size of the starting BaTiO₃. The inhomogeneous formation of necks among the particles and the inconsistent growth rate of neck were observed. This seems to result in a wide distribution of the particle size and the size of the pores among the particles. In the middle sintering period the agglomeration will proceed with disappearance of the voids in the cross-linked structure by adhesion among the particles. Eventually the compactness of the ceramics increases, however, the agglomerating region are quite different in size. In the final sintering period the local formations of abnormally grown particles were observed in the agglomerating region. The inhomogeneous mixing of large and small particles has the responsibility to form various pores having many different diameter and results in a decrease of the density. In conclusion the decrease in the density seems to be due inhomogeneous distribution of pore caused by the rapid rearrangement of the fine particles and the irregular formation of the necks among the particles.

View full abstract

Density and elastic moduli of aluminosilicate glasses of composition R₂O-and MO⋅Al₂O₃⋅rSiO₂ were measured, where R=Li or (1/2Li+1/2Na), M=(1/2Mg+1/2Ca), and r=2 to 8. The volume per one mol of tetrahedral cations (Al and Si) is constant in each series, implying that bulk modulus K varies in proportion to V_ip^-4/3 (V_ip: ion-pair volume). However, because of the introduced R⁺ or M²⁺ ions occupying the interstices of silica-like cages, the log-log plots of K vs. V_ip gave a line of slope-4. It is proposed that one may use the relation KV_ip⁴=constant as a reference which represents the constitution of such stuffed silica type network glasses, though not explainable by a model in favor of the predominant contribution of anion-anion repulsion energy to the lattice energy.

View full abstract